An open-source tool for generating representative Dutch distribution grid models using graph similarity measures and open data

Limited access to detailed distribution grid data can limit possibilities for research into the impacts of distributed energy resources and flexibility on Dutch electricity networks. This paper presents an open-source tool that generates realistic combined medium-voltage (MV) and low-voltage (LV) grid models using open available geographic datasets from Dutch distribution system operators (DSOs), archetypal neighborhood data, and open building registries. The tool extracts MV grids, assigns cable types based on detailed analysis of DSO data on actual networks, and matches LV grids to archetypal configurations using graph-matching techniques. Generated grids are formatted in the Energy System Description Language (ESDL) for co-simulation and analysis. Validation through load flow simulation shows the tools ability to extract and generate MV grid models that accurately reflect the real network behavior.